'Fallout 4' Benchmarks, And How To Disable VSync

[blazorthon]

It's not meant as a complete guide, just what they had.

Also, there's rarely any difference between an i5 and i7 of the same generation on a desktop CPU. Both have four cores, and the hyperthreading rarely makes a difference for gaming (on the i3 it would).

That depends on how well-threaded the game is and how well it scales across eight or more threads. The i7's Hyper-Threading is just as effective as it is on i3s with games that utilize more threads, such as BF3/4 multiplayer. If the game scales well on an FX-9590 as maddogfargo claims, then it is possible that Fallout 4 can take advantage of an i7, although I doubt it needs one regardless of the settings.

 

[jimmysmitty]

It's not meant as a complete guide, just what they had.

Also, there's rarely any difference between an i5 and i7 of the same generation on a desktop CPU. Both have four cores, and the hyperthreading rarely makes a difference for gaming (on the i3 it would).

That depends on how well-threaded the game is and how well it scales across eight or more threads. The i7's Hyper-Threading is just as effective as it is on i3s with games that utilize more threads, such as BF3/4 multiplayer. If the game scales well on an FX-9590 as maddogfargo claims, then it is possible that Fallout 4 can take advantage of an i7, although I doubt it needs one regardless of the settings.

I have been playing on my "currently stock" i5 4670K and a GTX 980Ti Strix and I also have the Steam FPS counter enabled and it has never dropped below 60FPS. I think an i5 is fine for this game.

 

[photonboy]

CPU scaling:
There's some CONFUSION that just because someone sees the CPU using eight cores of an FX-8350 (say roughly 30% each) that the game scales well across eight cores.

There's always a main thread of code for the game which can end up as the bottleneck in the system.

There is also something called "thread jumping" meaning the code can move between cores at certain times so it's difficult to determine how well "threaded" the game code actually is.

(You can use eight cores partially but not benefit more than say using THREE of them by disabling the other five. Rather than jumping between cores the thread execution just stays where it is or jumps a bit between the remaining cores... I think there's software to FORCE the main thread to stay on one physical core and to assign that core to its usage only to help eliminate bottlenecking.)

The ONLY way to properly determine how well game code is threaded is to run BENCHMARKS. First, to determine the main thread bottleneck (for example, by overclocking until no performance benefit is seen).

Secondly you can determine if threading helps which is a bit trickier. You can try setting the CPU frequency to the point at which the bottleneck disappears on the FX-8350 then start DISABLING cores. If you get to FOUR cores remaining and it works still then the remaining may not help... or try dropping the frequency a bit and experiment but it can be tricky outside of a pure benchmark program environment.

Anyway... a bit off topic but interesting. If I was an AMD CPU user I'd be definitely looking forward to DX12 gaming.

 

[hapkiman]

You say one of your systems is a "i7 5830k". Obviously a typo but was it was 5820k or a 5930k?

 

[davmazin]

The 2600k might be old but lets be honest there hasn't been any real improvement in cpu power for a few years now since its not a real big jump like from a q6600 to a 2600k wich was the jump i made it was like 60% to 70% and right now there isn't any cpu in the market that gives that big of a jump in cpu power, unless you overclock and really need something better for that i dont see the point in upgrading i suppose that maybe the 6xxx series might be worth but still i am not sure of any new cpus maybe if amd was a real competitor but as things stands its an intel monopoly even if amd is still there for me its just a facade

 

[splinter48708]

Not exactly a solution, but, more of a question: Why are AMD CPUs left out? This is a biased towards Intel article. To be fair, AMD CPUs are more than able to keep up with the more pricey Intel CPUs, IMHO, and should have been included.

 

[blazorthon]

Not exactly a solution, but, more of a question: Why are AMD CPUs left out? This is a biased towards Intel article. To be fair, AMD CPUs are more than able to keep up with the more pricey Intel CPUs, IMHO, and should have been included.

Again, they didn't go out and choose these systems to benchmark. These are just what they had on-hand. There is no bias in that.

 

[jimmysmitty]

Not exactly a solution, but, more of a question: Why are AMD CPUs left out? This is a biased towards Intel article. To be fair, AMD CPUs are more than able to keep up with the more pricey Intel CPUs, IMHO, and should have been included.

These are user systems. Meaning personal systems. This is not a review or benchmark done in the TomsHardware labs, where they have access to all the different hardware parts. These are the news teams personal systems so whatever they have is what they have.

 

[falchard]

On the Brightside, recording doesn't affect my frame rates using an AMD setup. With an FX processor and 290x GPU I have the same dip in FPS within the city regardless of if my session recorder is on or off.